Hi, there:
I noticed pheweb phenolist glob-files on this instruction page https://github.com/statgen/pheweb/blob/master/etc/detailed-loading-instructions.md#configuration-options. However, it won't work. It should be just "glob", not "glob-files".
The first time when I run pheweb process, I think pheweb automatically download the huge dbsnp reference files. If I already have these huge dbsnp files downloaded for another software, how could I speficy its location in the config.py file so that it won't need to be re-downloaded?
Previously, I have pointed out that many GWAS simply report A1/A2, not the strict REF/ALT. Once the reference genomes get bigger and bigger, the REF/ALT specification would become really indefinite and inconsistent. Therefore, I strongly suggest that pheweb would accommodate the use of A1/A2 and be able to map rsIDs with a better flexibility. Otherwise, most SNPs in a GWAS would have no mapped rsIDs after running pheweb. You guys only need to change one line (line 140), for add_rsids.py.
Before：
rsids = [rsid['rsid'] for rsid in rsid_group if cpra['ref'] == rsid['ref'] and are_match(cpra['alt'], rsid['alt'])]
After：
rsids = [rsid['rsid'] for rsid in rsid_group if (cpra['ref'] == rsid['ref'] and are_match(cpra['alt'], rsid['alt'])) or (cpra['ref'] == rsid['alt'] and are_match(cpra['alt'], rsid['ref']))]
Please kindly consider this.
Thank you & best regards,
Jie
Hi, there:
I noticed pheweb phenolist glob-files on this instruction page https://github.com/statgen/pheweb/blob/master/etc/detailed-loading-instructions.md#configuration-options. However, it won't work. It should be just "glob", not "glob-files".
The first time when I run pheweb process, I think pheweb automatically download the huge dbsnp reference files. If I already have these huge dbsnp files downloaded for another software, how could I speficy its location in the config.py file so that it won't need to be re-downloaded?
Previously, I have pointed out that many GWAS simply report A1/A2, not the strict REF/ALT. Once the reference genomes get bigger and bigger, the REF/ALT specification would become really indefinite and inconsistent. Therefore, I strongly suggest that pheweb would accommodate the use of A1/A2 and be able to map rsIDs with a better flexibility. Otherwise, most SNPs in a GWAS would have no mapped rsIDs after running pheweb. You guys only need to change one line (line 140), for add_rsids.py.
Before：
rsids = [rsid['rsid'] for rsid in rsid_group if cpra['ref'] == rsid['ref'] and are_match(cpra['alt'], rsid['alt'])]After：
rsids = [rsid['rsid'] for rsid in rsid_group if (cpra['ref'] == rsid['ref'] and are_match(cpra['alt'], rsid['alt'])) or (cpra['ref'] == rsid['alt'] and are_match(cpra['alt'], rsid['ref']))]Please kindly consider this.
Thank you & best regards,
Jie